模型42

Muse Wonder 登顶 Legora 法律基准

原标题:Legal work is a tough test for AI.

AI at Meta@AIatMeta发布时间 (北京时间)

法律工作是 AI 的一项严峻考验。 Muse Wonder 在 Legora 11 月法律工作 BAR 基准测试中获得最高综合质量分。在企業需要帮助的各类复杂任务上表现出色。

中文译文尚未完成,先为你展示原文。

Legal work is a tough test for AI.

Muse Wonder earned the highest overall quality score in Legora’s November BAR benchmark for legal work. A strong showing on the kinds of complex tasks businesses need help with.

The November BAR results are now available, showing how AI models compare on real-world legal tasks in Legora. A few highlights from this month’s results: → Muse Wonder ranks highest on overall quality and leads across four task types: analysis and answering, extraction, drafting, and editing. → GPT-6.2 Astra leads on comparison and review tasks, with overall quality 4% above November’s model average and cost per case 20% below average. → GPT-6.2 Sol delivers above-average quality at 78% below the average cost per case across models benchmarked in November.
查看原帖
Meta / Llama评测基准#评测/基准#Meta